Warrior General Games

lol awesome

Can you explain how the forumla works? In a way that people can understand, so when I post the poll they can decide if they want to vote using it or just the stats.

Originally posted by DrkSandman
Can you explain how the forumla works? In a way that people can understand, so when I post the poll they can decide if they want to vote using it or just the stats.

There are 4 factors: pancakes per offensive play, average team rushing yardage, hurries allowed per offensive play and sacks allowed per offensive play.

Each factor is then divided by the highest off all the teams for that factor. So the highest "factor" is 1. The pancake factor and the rushing factor are added, then the hurried and sacked factors are subtracted. So a "perfect" rating would be 2.0 (+1 for highest pancakes, +1 for highest rushing average, -0 for no hurries allowed and -0 for no sacks allowed).

Originally posted by cosine4
Originally posted by DrkSandman

Can you explain how the forumla works? In a way that people can understand, so when I post the poll they can decide if they want to vote using it or just the stats.

There are 4 factors: pancakes per offensive play, average team rushing yardage, hurries allowed per offensive play and sacks allowed per offensive play.

Each factor is then divided by the highest off all the teams for that factor. So the highest "factor" is 1. The pancake factor and the rushing factor are added, then the hurried and sacked factors are subtracted. So a "perfect" rating would be 2.0 (+1 for highest pancakes, +1 for highest rushing average, -0 for no hurries allowed and -0 for no sacks allowed).

From what I can tell, I think your formula favors rushing teams. After all, the best passing/worst rushing team will have 0 highest pancakes, 0 highest rushing average, -0 for no hurries, and -0 for no sacks, for a grand total of 0. On the other hand, the best rushing team/worst passing team will have +1 for highest pancakes, +1 for highest rushing average, -0 for no hurries, and -1 for tons of sacks, for a grand total of +1. This is because you can only get 1 hurry or sack on any given play, not both... and the worst teams won't give up so many hurries because they're so busy getting sacked. Now empirically there's no doubt that sacks and hurries will remain correlated, but I wouldn't be surprised to find that the team that gives up the most sacks is not the same as the team that gives up the most hurries.

I could be wrong, I couldn't figure out how to get at your formulas since I'm an excel n00b. However, I suspect I am not, because the teams with the most rushes this season are Fremont, ISB, and Tulsa.

I would suggest a similar formula, but with a single positive rushing factor (pancakes would suffice imo, but if not, then a weighted formula involving pancakes and rushing yards), and a single negative passing factor (a weighted formula of sacks and hurries, with sacks weighed more heavily of course). This way the passing and rushing games are given equal weight, and the best O-line will be the one who is best at doing both.

The stats aren't completely normalized, meaning the highest will always be 1 but the lowest won't necessarily be 0. The zero point would be (zero pancakes) and the worst rushing team got some pancakes. But the way it was calculated still did favor rushing teams some. I adjusted the calculation so that it the passing factors are per passing attempt and the rushing factors are per rushing attempt.

Oh yeah, the above link has been updated with the new file.

Originally posted by cosine4
The stats aren't completely normalized, meaning the highest will always be 1 but the lowest won't necessarily be 0. The zero point would be (zero pancakes) and the worst rushing team got some pancakes. But the way it was calculated still did favor rushing teams some. I adjusted the calculation so that it the passing factors are per passing attempt and the rushing factors are per rushing attempt.

not trying to be a pain - thanks, I'll leave you alone now

Originally posted by tjking82
Originally posted by cosine4

The stats aren't completely normalized, meaning the highest will always be 1 but the lowest won't necessarily be 0. The zero point would be (zero pancakes) and the worst rushing team got some pancakes. But the way it was calculated still did favor rushing teams some. I adjusted the calculation so that it the passing factors are per passing attempt and the rushing factors are per rushing attempt.

not trying to be a pain - thanks, I'll leave you alone now

So, is his new forumla better iyo? I think it sounds pretty good.

Originally posted by DrkSandman
Originally posted by tjking82

Originally posted by cosine4


The stats aren't completely normalized, meaning the highest will always be 1 but the lowest won't necessarily be 0. The zero point would be (zero pancakes) and the worst rushing team got some pancakes. But the way it was calculated still did favor rushing teams some. I adjusted the calculation so that it the passing factors are per passing attempt and the rushing factors are per rushing attempt.

not trying to be a pain - thanks, I'll leave you alone now

So, is his new forumla better iyo? I think it sounds pretty good.

Yeah I think it is much better. It looks fairly balanced now

Good spreadsheet, but the rush attempts column wasn't sorted to match up with the proper teams, and the results are currently skewed.

After looking at this again, you may want to add kick and punt returns to the rush plays, as those also allow opportunities for pancakes. If you don't do that, passing teams are going to come out far ahead of rushing teams on the pancake ratio. That's why the Tulsa Teamsters look so good now, as they have 917 pancakes against 98 rush plays (which is actually French Licks' total).

I'll take a look at it. I thought I sorted everything the same way, but must have missed some.

Originally posted by cliffhangergg
Good spreadsheet, but the rush attempts column wasn't sorted to match up with the proper teams, and the results are currently skewed.

After looking at this again, you may want to add kick and punt returns to the rush plays, as those also allow opportunities for pancakes. If you don't do that, passing teams are going to come out far ahead of rushing teams on the pancake ratio. That's why the Tulsa Teamsters look so good now, as they have 917 pancakes against 98 rush plays (which is actually French Licks' total).

We had 98 rush plays? Already? I was hoping to stay under 100 for the season

^ Ok, now it is correct.

Originally posted by cosine4

http://www.fisherrobotics.com/oline.xls

Originally posted by cosine4
^ Ok, now it is correct.

Originally posted by cosine4


http://www.fisherrobotics.com/oline.xls



oh geez now I look really bad

^ How so?